Development of a Production Line Appropriate Test for

Development of a Production Line Appropriate Test for Teflon Adhesion on 304 Stainless Steel 0.01-0.03″ Diameter Coronary Guide Wires

A Senior Project

presented to

the Faculty of the Materials Engineering Department

California Polytechnic State University, San Luis Obispo

In Partial Fulfillment

of the Requirements for the Degree

Bachelor of Science

by

Danielle Dunham

June, 2013

© 2013 Danielle Dunham

i

Approval Page

Project Title: Development of a Production Line Appropriate Test for Teflon Adhesion on 304 Stainless Steel 0.01-0.03″ Diameter Coronary Guide Wires

Author: Danielle Dunham

Date Submitted: June 7, 2013

CAL POLY STATE UNIVERSITY

Materials Engineering Department

Since this project is a result of a class assignment, it has been graded and accepted as fulfillment of the course requirements. Acceptance does not imply technical accuracy or reliability. Any use of the information in this report, including numerical data, is done at the risk of the user. These risks may include catastrophic failure of the device or infringement of patent or copyright laws. The students, faculty, and staff of Cal Poly State University, San Luis Obispo cannot be held liable for any misuse of the project.

Prof. Blair London ____________________________

Faculty Advisor Signature

Prof. Richard Savage ____________________________

Department Chair Signature

ii

Acknowledgements

I would like to thank Abbott Vascular for sponsoring the project, providing wires for testing and

manufacturing the prototypes. Specifically I would like to thank Matt Gillick for his help designing and

manufacturing the prototype, John Simpson for his help and advice throughout all areas of the project,

and Jeff Dooley for his help with data analysis. I would also like to thank Dr. Blair London for his help and

guidance throughout the project and Daniel DeBruler for moral support.

iii

Table of Contents Introduction ............................................................................................................................................ 1

Coronary Artery Disease ...................................................................................................................... 1

Coronary Angioplasty .......................................................................................................................... 2

Guide Wires ......................................................................................................................................... 2

Teflon .................................................................................................................................................. 3

Adhesion of Teflon ........................................................................................................................... 4

Current Technology ............................................................................................................................. 4

Pull Off Tests.................................................................................................................................... 4

Peel Tests ........................................................................................................................................ 5

Scratch Tests.................................................................................................................................... 6

Materials and Equipment Used for Testing .............................................................................................. 6

Design 1 .................................................................................................................................................. 7

Adhesion Testing ................................................................................................................................. 7

Design2 ................................................................................................................................................... 8

Coating Failure .................................................................................................................................... 8

Adhesion Testing ................................................................................................................................. 9

Design 3 ................................................................................................................................................ 10

Adhesion Testing ............................................................................................................................... 11

Design 4 ................................................................................................................................................ 12

Prototype Manufacture ......................................................................................................................... 12

Equipment Used for Testing............................................................................................................... 12

Realistic Constraints .......................................................................................................................... 12

Prototype Design 1 ................................................................................................................................ 13

Adhesion Testing- Method 1 .............................................................................................................. 14








iv

Process Improvement ............................................................................................................................ 21


Failure Morphology ....................................................................................................................... 22

Electrical Continuity ....................................................................................................................... 23

Spray Coated Wires ............................................................................................................................... 24


Conclusions and Recommendations ...................................................................................................... 26

References ............................................................................................................................................ 27

v

List of Figures Figure 1- Healthy vs. Disease artery ......................................................................................................... 1

Figure 2- Coronary Angioplasty ................................................................................................................ 2

Figure 3- Components of a coronary guide wire ....................................................................................... 3

Figure 4- The structure of Teflon. ............................................................................................................ 3

Figure 5- A schematic of a pull off test fixture. ......................................................................................... 5

Figure 6- An example of a 180° peel test. ................................................................................................. 5

Figure 7- An apparatus for a scratch adhesion test................................................................................... 6

Figure 8- The two methods of producing a shear force: twisting and sliding. ........................................... 7

Figure 9- The first testing apparatus. ....................................................................................................... 8

Figure 10- Effect of the contact cylinder on the coating. .......................................................................... 9

Figure 11- Poorly adhered wire vs. well adhered wire failure morphologies ............................................. 9

Figure 12- An apparatus to test a horizontal lever bar design. ................................................................ 11

Figure 13- Solidworks assembly of Prototype Design 1........................................................................... 13

Figure 14- Determination of weight increments ..................................................................................... 14

Figure 15- Results from consistent increment test method. ................................................................... 15

Figure 16- Prototype Design 2 with weight attachment and wire clamp additions. ................................ 16

Figure 17- Results from testing with Prototype Design 2. ....................................................................... 17

Figure 18- Results from test for process drift. ........................................................................................ 17

Figure 19- Results from testing with electrical continuity. ...................................................................... 19

Figure 20- Results from testing with electrical continuity with a fast slide speed. ................................... 20

Figure 21- Results from testing with electrical continuity with a slow slide speed .................................. 21

Figure 22- Failure morphologies at each processing step. ...................................................................... 22

Figure 23- A comparison of the well adhered wires after each processing step. ..................................... 23

Figure 24- Determination of process drift . ............................................................................................ 24

Figure 25- Failure morphology of the spray coated wire at 1.25 lb.. ....................................................... 25

Figure 26- Results from testing spray coated wires. ............................................................................... 26

List of Tables

Table I- Comparison of Results from two Potential Test Methods….………………………………………………………11

vi

Abstract

A test was developed to determine relative adhesion of Teflon-coated 304 stainless steel guide wires.

Two sets of wires were provided by Abbott Vascular for testing: one set was known to have poor Teflon

adhesion, the other had acceptable adhesion. A prototype device was designed and manufactured that

produced a shear force on the coating by dragging the wire under a ½ inch-diameter 440C hardened

stainless steel cylinder. Testing consisted of increasing the force on the wire during each consecutive

pass with the cylinder until the Teflon coating failed. The force on the wire during the failure producing

pass was recorded as the failure weight. Failure was determined using a low magnification microscope.

When the Teflon had been removed and the bare wire was visible, the coating had failed. The prototype

was successful in showing a difference in adhesion between the two sets of wires; the test produced a

higher average failure weight for the well adhered wires than the poorly adhered wires. An independent

sample t-test was conducted to compare the results of the prototype after testing poorly adhered and

well adhered wires. There was a statistically significant difference between the two distributions with a

p-value less than 0.001. The test was altered to include electrical continuity as a more objective method

of detecting failure. The test was successful in using electrical continuity to determine a statistical

difference between the two adhesion levels. The latest test method and prototype were then used to

test samples of one batch of wire at three different processing steps: as-coated, as-straightened, and

post burn-in. All three processing steps were statistically different; straightening caused a decrease in

adhesion levels, while burn-in caused a slight increase in adhesion.

Keywords: Materials Engineering, Adhesion Testing, Teflon, Coronary Guide Wires

1

Introduction Coronary guide wires are often produced with a Teflon coating to reduce friction. Good

adhesion between the stainless steel wire and the Teflon coating is important because guide wires with

flaking Teflon are dangerous to patients. Currently, there is no quantitative test available to determine if

the adhesion of the Teflon to the steel is adequate near the beginning of guide wire processing in a

production line setting. Pull off and peel off tests are not feasible for small wires, and scratch tests are

not easily accomplished in a production line setting. Flaking of the Teflon coating becomes visually

apparent late in the processing of the wires, and the wires must be disposed of after value has been

added. An adhesion test that is quick and easy to perform on thin wires is necessary to determine

adhesion of Teflon to the steel early in processing to reduce the value of wires scrapped for inadequate

Teflon adhesion. In addition, the test could be used for process improvement, comparison of different

coating methods, and maintaining consistent adhesion levels between facilities.

Coronary Artery Disease

Coronary Artery Disease is a condition caused by an excess of fat and cholesterol in the coronary

arteries. When cholesterol is digested in the intestines it is transported by lipoproteins. After digestion,

the lipoproteins enter the liver and pick up fat and cholesterol to become Low Density Lipoproteins

(LDLs). The LDLs move through the arteries and deposit fat and cholesterol. When there is too much

cholesterol being deposited, plaque begins to build up along the walls and block off the arteries leading

to coronary artery disease (Figure 1).

Figure 1- Examples of a healthy artery on the left and an artery on the right clogged with excess fat and cholesterol limiting blood flow1.

2

Narrowing of the coronary arteries prevents blood from reaching the heart and can lead to a

heart attack2. There are a range of treatments depending on the severity of the condition. The symptoms

can be relieved through medications designed to reduce blood pressure and cholesterol; life style

changes are also recommended including a decrease in consumption of high cholesterol, high fat foods

and quitting smoking3. For extreme cases surgery, or coronary angioplasty, is necessary to prevent a

heart attack.

Coronary Angioplasty

During coronary angioplasty a balloon catheter is used to push plaque against the walls of the

artery, opening the artery and increasing blood flow (Figure 2). The femoral artery in the leg is most

commonly used for entry, though the brachial artery in the arm can also be used. A guide catheter is

inserted into the selected artery and guided to the obstruction. A coronary guide wire with a radio

opaque tip is then inserted into the guide catheter. Using x-rays the guide wire is guided past the

obstruction and a deflated balloon catheter is placed over the guide wire. The balloon is inflated once it

reaches the blockage. The balloon catheter, guide wire, and guide catheter are then removed5.

Figure 2- A diagram of an inflated balloon catheter. The balloon catheter is threaded over the guide wire through the obstruction and inflated to open up the artery by pushing plaque against the walls4.

Guide Wires

Coronary guide wires have three main components: the core, distal tip, and coating (Figure 3).

The core is a stainless steel or nitinol wire that runs along the length of the guide wire. The material and

the diameter of the core can vary to alter the stiffness, steering, support, and torque of the wire. The

3

core usually tapers down toward the distal tip, though the amount and type of tapering will vary. The tip

can be an extension of the core, or a two piece design where metal will bridge the gap between the tip

and the core. The tip design can alter the torque control, flexibility and stiffness of the guide wire.

Coatings are used to reduce friction and help the equipment such as a balloon catheter to easily slide

along the wire. The coating on the core also controls the tactile feel of the guide wires6.

Figure 3- Components of a coronary guide wire. A two piece tip design is shown with a coil around the tip. The coating is along the core and stops at the tip. Typical wire diameters range from 0.01 to 0.03 inches.

Teflon

Polytetrafluoroethylene (Teflon) is a fluoropolymer (Figure 4) that was first discovered

accidentally in 1938 by Roy J. Plunkett while Du Pont and GM were working together to produce an

efficient refrigerant8. Teflon was discovered to have a number of beneficial properties including

insolubility in a number of different common solvents, high hydrophobicity, and a high melting

temperature. While the initial production of Teflon was accidental, the unique properties drove Du Pont

to continue developing PTFE for military and finally commercial use9.

Figure 4- The structure of Teflon. Teflon is a fluoropolymer with an amorphous structure of fluorine saturated carbon chains8.

0.01″ - 0.03″

4

Teflon is highly hydrophobic due to a low surface energy10. The low surface energy causes water

to water interactions to be stronger than water to Teflon interactions. Water will bead up on the surface

of hydrophobic materials rather than wetting the surface. The hydrophobicity of Teflon reduces the

friction on the surface of a substrate; coating guide wires with Teflon reduces the sliding friction in the

arteries by up to 50%11. Hydrophilic coatings have a higher surface energy that attracts water molecules.

The hydrophilic coatings will become slippery when placed in contact with blood or saline solution and

can decrease the sliding friction to 16% of the original friction in the artery11. Lower sliding friction

allows the wire to move easier through the artery; however, less friction also increases the risk of the

wire slipping and perforating the artery. The higher friction of the Teflon coating compared to

hydrophilic coatings produces a wire that is more controllable by the doctor, but still moves easier

through the arteries than uncoated steel6.

Adhesion of Teflon

Low surface energy prevents Teflon was sticking or bonding to most surfaces, therefore

intentional adhesion to steel guide wires is difficult to achieve. In general, substrates can be altered

either chemically or physically to encourage adhesion of difficult coatings. Chemically etching the

surface adds reactivity to the surface to encourage chemical bonding. Teflon has a low reactivity due to

the fluorine atoms completely saturating the carbon chain; therefore, Teflon will not bond chemically to

other surfaces. Instead, physical etching of the surface is often used to increase the surface roughness

and encourage physical bonding through mechanical interlocking10.

Current Technology

There are currently several methods for determining coating adhesion including pull off, peel,

and scratch tests. Each test has specific constraints for sample preparation and geometries. The results

of each test may not be comparable to one another due to a difference in the properties obtained such

as shear strength versus tensile strength.

Pull Off Tests

Pull off tests measure the normal force required for a coating to fail. A loading fixture is secured

to the coating with an adhesive, and the coating is subjected to a force in tension until the coating fails

(Figure 5). Adhesion is analogous to the pull off strength, which is determined using the maximum load

and the coating’s surface area that is subjected to the load12. The coated samples must have a flat area

wide enough to allow the apparatus to attach without producing faulty data due to edge effects. Also,

5

the substrate should be rigid enough to support the tensile force; therefore, pull off tests are not

suitable to measure adhesion of coated wires.

Figure 5- A schematic of a pull off test fixture. The adhesive attaches the coating to the loading fixture while the central grip is used to apply tension until the coating is removed12.

Peel Tests

Peel tests measure adhesion as the peel or stripping strength of

the sample. There are several angles that can be used in peel tests such as

T shape or 180°. In all types of peel tests the testing specimen must be

prepared so that one end is securely adhered and the other end has the

coating and substrate separate and unattached. The substrate will be

secured in the apparatus with the open end of the coating clamped. In a

180° peel test the coating is pulled at a separation angle of 180° until half

of the bonded area is removed (Figure 6)13. The peel strength is recorded

as the average load necessary to peel off the coating. Because peel tests

require a specific test specimen to be produced with only half of the

coating securely adhered, the test is not acceptable for a production line

adhesion test. Additionally peel tests are not able to be done on small

diameter wires; therefore the test is not acceptable for coronary guide wires. Figure 6- An example of a 180° peel test. The coating is pulled straight down while the substrate is held in place by the grip13.

6

Scratch Tests

Scratch tests measure adhesion as the force required to scratch the coating off of the substrate.

Scratch tests use a weight to apply a constant force on a steel loop that scratches the surface of the

coating (Figure 7). A starting force is used to scratch the coating; if the coating is removed a lighter force

is used until the coating is no longer removed. If the coating was not removed from the initial force then

the force is increased until the coating is removed. The adhesion calculation is based on the minimum

force required to remove the coating14. Scratch tests must have level planar test specimens; the test will

not be accurate with cylindrical wires and would be difficult to conduct on small diameter wires.

Figure 7- An apparatus for a scratch adhesion test. The coated substrate is placed under the metal loop. The platform is pushed away from the operator causing the loop to scratch the coating14.

Materials and Equipment Used for Testing Two lots of wires were provided by Abbott Vascular for testing of the prototype design. Both

wires were 0.013” 304 stainless steel covered with a green-pigmented Teflon coating. One wire lot was

known to have a poorly adhered Teflon coating and had been through all processing steps, while the

other lot had an acceptable Teflon coating but was in an as-coated condition. A successful prototype will

be able to show a statistically significant difference between the poorly adhered and acceptable (well

adhered) wires.

An adhesion test typically involves identifying failure of the coating; however, failure of the thin

wire coating is unable to be seen by the naked eye. Instead a low magnification USB microscope

Steel

Loop

Weight to

increase

force

7

(Zarbecco Miscope) was used to better image the coating. The wire was focused at about 50X and 130X,

and the image was displayed on a laptop computer.

Design 1 A scratch test was used as the template for the overall design. Because the stylus of a scratch

test would be too difficult to center on a 0.01” diameter wire quickly and consistently, a bulk shape was

used to produce a similar type of shear force on the coating. A square block was first used to test the

ability of a bulk object to produce an appropriate shear force on the wire to obtain failure. The block

was moved along the wire, producing a shear force on the coating until the Teflon coating failed.

Adhesion Testing

Early test runs revealed there are two ways to produce a shear force on the coating capable of

removing the Teflon: sliding the block along the section of coated wire, or placing the block in one place

and twisting the block. Failure was defined as removing the Teflon coating, which can be seen clearly

when the brown oxide or bare wire has been uncovered. Twisting the block proved to destroy the

coating more significantly than sliding the block along the section of wire when the same size weight

was used (Figure 8).

Figure 8- Two methods of producing shear force along the wire: sliding the block and twisting the block. The poorly adhered wire after a 0.68 lb block was slid across (a) and the poorly adhered wire after the same block was twisted in place wire causing the Teflon to fail (b).

0.01 in

0.01 in (a) (b)

Torn Teflon

Coating Steel Wire

Teflon coated wire

8

Design2 A cylinder contact shape replaced the square block because the cylinder involves less

maintenance than the sharp edge of the square block. In addition, a cylinder can be rotated as it begins

to wear to increase the lifetime of the cylinder. A cylinder does not lend itself to a twisting method;

therefore the method of sliding the cylinder along the wire was adapted for further testing.

A temporary apparatus was constructed of pieces of available metal to test the ability of the

method to cause failure consistently before resources were put into building a prototype (Figure 9). The

apparatus consisted of a lever bar that held the cylinder in position. The wire was taped to a base plate

and the cylinder was rested on the wire. The weights were stacked on the close end of the lever bar to

add additional force to the cylinder as it contacted the coating. The base plate was then slid away from

the operator, causing the cylinder to put a shear force on the Teflon coating.

Figure 9- The first apparatus built to test the concept. The wire is taped to the base plate. The contact cylinder puts a shear force on the wire coating as the base plate is pushed away from the operator. The weights are added on top of the lever bar to increase the force.

Coating Failure

Wire that has been tested with the cylinder but has not failed tended to develop a dark line

down the center of the wire (Figure 10). The line down the center was not considered failure; failure

was considered to be when the bare wire or brown oxide was visible down the center of the wire (Figure

11). Well adhered and poorly adhered wires failed slightly differently in that the poorly adhered wire

Support Pillars

Lever Bar

Contact cylinder

Teflon

Coated

Wire

Weight

s

Base Plate

9

had ragged edges where the poor adhesion allowed the Teflon to be pushed aside during testing, while

the well adhered wire showed straight, consistent edges (Figure 11).

Figure 10- An untested wire (a) versus a wire that has been rubbed with the contact cylinder (b). After the contact cylinder made contact and slid across the coating, a line formed down the center of the wire.

Figure 11- A failed poorly adhered wire (a) versus a failed well adhered wire (b). The poorly adhered wire failed with more ragged edges; while the well adhered had straight consistent edges.

Adhesion Testing

Two test methods of testing were explored with the apparatus. The first involved using a single

weight and counting the number of passes until failure of the Teflon coating. The wire was taped down

and checked for defects. The chosen weight was set on the cylinder, the cylinder was set on the plate,

0.01 in

0.01 in

0.01 in

0.01 in

Teflon Coated

Wire

(a) (b)

Aluminum

Mount

(a) (b)

Stainless Steel

Wire

10

and the plate was pushed away from the user to produce a shear force. The wire was then checked for

failure. If no failure was found the cylinder was again set down on the wire with the same weight and

the plate was pushed away from the user. The cycle continued until failure occurred, and the number of

passes required to cause failure were recorded.

The second method involved increasing the weight after each pass. The same steps were

followed until the first pass was checked for failure. Then the weight was increased on the second pass,

and the wire was checked for failure. The increase in weight on each pass continued until failure

occurred, and the weight required to cause failure was recorded. Both methods showed a difference

between the poorly adhered and well adhered wire adhesion levels (Table I). The first method was

accomplished using a constant 1.29 lb weight. The second method began with a 1.29 lb weight. If the

wire did not fail the weight was increased to 1.49 lb, 1.86 lb, 2.07 lb, and 2.13 lb until failure.

Table I. Comparison of Results from the Two Potential Test Methods

Design 3 The second method was chosen for further designs because it could accommodate a wider

range of adhesion levels. A constant weight chosen for Method 1 that worked with a well adhered

coating may be too high to accommodate a poorly adhered coating: the coating would fail on the first

pass producing attribute rather than variable data. Conversely, a weight that worked well with a poorly

adhered coating may take far too many passes for a well adhered coating; the test would be

unacceptably long for a production line. The second method allowed a low weight to be used in the

beginning of testing, accommodating the poorly adhered wire by not causing failure on the first pass. By

increasing the weight on each pass the method also accommodated the well adhered wire without

requiring an excessive number of passes to failure.

Adhesion Level

Method 1: # of Passes to Failure

Method 2: Failure Weight (lb)

Poor 1 4 1.86

Poor 2 4 1.86

Good 1 6 2.13

Good 2 7 2.07

11

A design change was also implemented for Design 3. Rather than stacking additional weights on

top of the bar for each pass, the new design involved a horizontal bar with a single weight (Figure 12).

The weight started at the far end of the lever bar, placing the minimum force possible on the wire

during the first pass. For each consecutive pass the weight was moved forward on the bar, increasing

the force on the wire. The method was considered more user-friendly because the operator would not

have to work with multiple weights. In addition, 440C hardened stainless steel replaced the metal used

for the contact cylinder in the previous design.

Figure 12- The apparatus was modified to include a horizontal lever design. A free weight was placed on the lever bar at the positions marked. Moving the weight forward on the bar increased the force on the wire. The cylinder used was 440C hardened stainless steel.

Adhesion Testing

The apparatus was tested using the increasing weight method. Marker lines designated where

the weight should be placed during each consecutive pass of the cylinder. The failure weights with the

apparatus are unknown because a method of measuring the force on the wire had not been created yet.

Instead, the positions of the weight were recorded for each wire. The farthest back the weight could be

placed on the lever bar was 2.25″ from the far side of the lever. The position for the second pass was

2.5″, and the weight moved forward 0.5″ on each successive pass. The testing was completed with a

1.75 lb weight, the heaviest weight available at the time. The poorly adhered wire failed at 4.5″ from the

far end of the lever. The apparatus was unable to cause failure on the well adhered wire with the

440C Stainless

Steel contact

cylinder

Base Plate

Sample

Wire

Movement of

weight along the bar

to increase force on

the wire

12

weights available; however, the test was successful in showing a clear difference between the adhesion

levels of the poorly adhered and well adhered wires.

Design 4 For the next design, the contact cylinder was moved back so that the lever bar extended past

the contact cylinder. The extra leverage allowed a higher force to be applied to the well adhered wire

without increasing the weight. An accurate 2 lb weight was obtained for the design and the base plate

was replaced with a linear slide (Del-Tron S1-3 ball slide) to decrease friction and maintain a controlled

motion. However the test was unable to be run because the prototype was not stable enough to apply

the 2 lb weight to the wire.

Prototype Manufacture

Equipment Used for Testing

A method for measuring the force on the wire was necessary to continue using Design 4. A

Denver Instrument scale was used. The scale was placed on a table with a copper wire secured to a hook

underneath the scale. The wire was then secured to the contact cylinder and the force was recorded for

several positions of the weight along the lever bar. A calibration curve (force on the wire versus distance

along the lever bar) was plotted and used to determine locations of the weight for specific down forces

on the wire.

Realistic Constraints

The project was constrained in multiple ways, mostly by manufacturability and economic

constraints. Several design decisions were made to produce a working prototype that was easy to

manufacture. The linear slide was used as a pre-manufactured part. The slide satisfied the need for a

low friction sliding surface, but did not necessitate extra time to manufacture. When the wire clamp was

added several parts needed to be altered to accommodate the extra height. The clamp could have been

made thin enough to prevent the need for changing the other already manufactured parts; however,

such a thin metal plate would likely bow during manufacture. Therefore the plate was made to be 1/8

inch thick, and the pillars had to be remanufactured.

Design decisions were also influenced by economic constraints. The developed test needed to be quick

and easy to perform in the production line. An atomic force microscope (AFM) adhesion test could be

Lever Bar

13

conducted on the wires; however, additional training and time to complete the test would increase the

cost of testing. Abbott was unlikely to use any test that needed large amounts of training or time to

complete. Also, a complete weight set would have been ideal during early testing to determine the best

weight to use for subsequent testing; however, the sets were expensive. Therefore two pounds was

decided on based solely on available weights, rather than with experiments to determine the most

acceptable weight for the test.

Prototype Design 1 The first prototype was modeled after Design 4 (Figure 13). The entire prototype was screwed

onto a base plate. The base plate, pillars, lever bar, and cylinder clamp were manufactured from 6061-

T6 aluminum. The contact cylinder was the same 440C hardened stainless steel, and the pivot bar was

also stainless steel. Prototype Design 1 did not include a way to clamp the wire in place, or an

attachment for the weight. The wire was taped in place during testing and marker lines were placed on

the lever bar to indicate where the 2 lb free weight would be placed during testing.

Figure 13- Solidworks drawing of Prototype Design 1. The linear slide and contact cylinder were used from Design 4, as was the overall design. The wire was taped to the linear slide and a 2 lb free weight was used on top of the lever bar at differing locations.

Base Plate

Linear Slide

Contact

Cylinder

Lever Bar

Pivot Bar

Wire Sample

Marker locations

for weight

14

Adhesion Testing- Method 1

The necessary weights and increments to produce failure in the two types of wires were

unknown when Prototype Design 1 was constructed. The first testing method was an attempt to gain

insight on the necessary increments. The first position was the lever bar with no additional weight: 0.42

lb. After each consecutive pass the weight was increased to 1.02, 1.7, 2.05, 2.3, and 2.5 lb. The test was

run on 10 samples of each wire (Figure 14). All of the poorly adhered wires failed on the third pass, or

1.7 lb; the well adhered wires failed on the fourth through sixth passes, indicating a difference in

adhesion levels of the two wire lots.

Figure 14- Results from 10 test runs on each wire set using Prototype Design 1. The increments were chosen at random starting at 0.42 lb and 1.02 lb. No wires failed until 1.7 lb, where all of the poorly adhered wires failed.


All of the poorly adhered samples failing at 1.7 lb implied that the coating could have failed at

lower weights if the increments were finer. In addition, a standard deviation of 0 makes the data difficult

to analyze statistically; T-tests cannot be run on populations with no standard deviation. Finer

increments could produce a more realistic distribution for the poorly adhered wires. Therefore, the next

step was to reduce the increments. The test started at 1 lb for the first pass, and the increments were

Poorly Adhered S = 0 X = 1.7 Well Adhered S = 0.12 X = 2.29

15

reduced to 0.1 lbs for the next test. Twenty wires were tested from each adhesion type, or forty wires

total (Figure 15).

The poorly adhered wire had a failure weight distribution with a non-zero standard deviation

with the new method. The average failure of the poorly adhered wire was 1.31 lb, and the well adhered

wire failure average was 1.85 lb. An independent T-test was performed on the two sample distributions

and the p value was less than 0.001; the populations were statistically different.

Figure 15- Results from testing 20 wires of each adhesion type with the new test method: 0.1 lb increments starting at 1 lb. The poorly adhered wire had a nonzero standard deviation, indicating a successful change in testing method.

Prototype Design 2 After initial testing showed that the prototype was capable of determining a statistically

significant difference between the two adhesion types, a wire clamp and weight attachment were added

to the prototype (Figure 16). The wire clamp ensured that the wire was securely attached during every

test run in the same position. The clamp consisted of a metal plate that screwed into the linear slide.

There was a groove down the center of the plate to locate the wire, and a screw and washer system on

either side to hold the wire in place during testing. The weight attachment provided a more consistent

method of producing force on the wire, reducing user error. The weight attachment had ball plungers on

Poorly Adhered S = 0.05 X = 1.31 Well Adhered S = 0.14 X = 1.85

16

either side that locked into detents on the side of the lever bar. The detents were placed so that the

lowest setting on the prototype was 0.95 lb, and the increments were 0.1 lb for each consecutive pass.

Figure 16- Prototype Design 2, an improvement over Prototype Design 1 including weight attachment and wire clamp additions. The weight is held in place by ball plungers. The wire uses a groove to locate the wire and a washer and screw on each side to hold the wire in place.

Adhesion Testing

Twenty tests were run on each adhesion type with the new prototype additions (Figure 17). The

poorly adhered wire had an average of 1.28 lb which is consistent with the previous test. The well

adhered wire failed at an average of 1.75 lb; which is 0.1 lb lower than the previous test. The difference

in values could be due to user variance. The edges (the beginning and the end of the test run) were

checked for failure even though the areas had added variables: if the cylinder was set down on the wire

harder than previously, the wire would be likely to fail under less weight. Also, if the linear slide was run

into the screws at the end of the test run the slide would bounce back, producing an additional force on

the wire. The difference could also be due to the operator’s inability to maintain a consistent

determination of failure with the microscope.

Ball Plunger

Locating Groove

Washer/Screw

System

Detents

Screws holding

down wire clamp

17

Figure 17- 20 test runs were performed on each wire set using Prototype Design 2. The average of the well adhered wires dropped 0.1 lb, though the poorly adhered wire data remained consistent. A test was also completed to determine if there was process drift. If there was a build-up of

Teflon on the wire, or if the contact cylinder began to wear, the values could progressively increase or

decrease over time. To test for any potential drift, 40 wires were tested in a random order. The residuals

were then plotted against the testing order (Figure 18). There was no clear increase or decrease in

residuals; however, the greatest 3 residuals were recorded in the last 11 data points. Therefore

additional testing would be advised to determine if the cylinder should be rotated after 40 test runs.

Figure 18- Test for process drift. The wires were tested in a random order and each residual was calculated by comparing the data point with its own population mean.


18

Prototype Design 3 The variability due to operator variance was addressed in the third prototype design. One way

to reduce operator variance was to use a more objective method of failure detection. Electrical

continuity replaced the low magnification microscope for failure detection. The wire clamp was

electrically isolated from the prototype by removing the screws holding the clamp in place (Figure 16)

and replacing them with double sided tape. Teflon was removed from one side of the wire and one lead

of a voltmeter was attached to the bare metal. The other lead was attached to the contact cylinder.

When the cylinder contacted on the intact-coating wire the circuit would be open and there was no

continuity. The test was run using the same method as before: increasing the weight after every pass.

Once the cylinder broke through the coating and contacted bare metal the circuit would be closed, the

voltmeter would register an electrical continuity, and the failure weight would be recorded.


Twenty wires of each adhesion type were tested using electrical continuity as a method of

failure detection (Figure 19). The microscope was used periodically to determine how the continuity

method agreed with the previous method of failure detection. The sample distributions were still

significantly different with a p-value of <0.001; however, the averages of both adhesion types were

increased about 0.3 lb. The microscope verified that the wires were failing much later than previous

tests. The well adhered wires were unable to show a clear failure with the microscope: at the end of the

range of the prototype the well adhered wires showed a thin line of oxide that was not consistent with

previous images for failure where the bare wire was visible. The voltmeter showed electrical continuity

when the oxide line appeared; however, if the microscope had been used the wires would not have

failed. The increase in failure weight using both methods of detection implied that something on the

prototype was altered to obtain such different results.

19

Figure 19- Results from testing twenty wires of each adhesion type with electrical continuity as failure detection. The average failure weights of both adhesion types increased by 0.3 lb, confirmed with the microscope.


A possible reason for the shift of the distributions was the removal of the screws that held down

the wire clamp (Figure 16); the movement of the slide was altered by removing the screws. When the

screws were present the slide was moved forward quickly; however, the slide had to be slowed to

prevent the cylinder from running into the screws and bouncing back, doubling the force on the end of

the wire. When the screws were removed the slide could be moved forward at a relatively constant

speed from start to finish. The change in speed introduced by the screws may account for the difference

in failure weights.

The screws were replaced to maintain a similar slide speed as prior tests and 20 wires of each

adhesion type were tested, using electrical continuity to detect failure (Figure 20). The sample

distributions were still statistically different with a p-value of <0.001. The averages of both distributions

were increased by 0.1 lb when compared to the previous tests that used the microscope as a method of

failure detection.

2.652.452.252.051.851.651.45

7

6

5

4

3

2

1

0

Failure Weight (lb)

Fre

qu

en

cy

Poorly Adhered

Well Adhered

Variable


20

Figure 20- Results from twenty tests run on both adhesion types using electrical continuity as a method of failure detection. The screws were replaced in this method to maintain all controllable variables.


The increase in failure weights when using the electrical continuity as a method for failure

detection was likely due to the fast motion of the slide. When the slide was moved forward quickly, as it

had been in all previous testing, the voltmeter was unable to detect failures right when they occurred.

Therefore, the next change was to slow down the speed of the slide. The slide was moved forward at a

speed that required 3-4 seconds to complete the motion, rather than 1 second in previous testing.

Twenty wires were tested using electrical continuity to detect failure and moving the slide at a

slower speed (Figure 21). The averages of both adhesion types were the same as the test with a fast

slide, 1.47 lb for the poorly adhered wire and 1.91 lb for the well adhered wire. Because the voltmeter

was now able to detect failure as it occurred, the increase in average failure weights was likely due to

the speed of the slide altering the results. The increase in average failure weights indicated that the

speed of the slide was an important variable that should be regulated.

2.352.151.951.751.551.35

7

6

5

4

3

2

1

0

Failure Weight

Fre

qu

en

cy

Poorly Adhered

Well Adhered

Variable


21

Figure 21- Results from slowing the speed of the slide to about ¼ the previous speed and testing 20 wires of each adhesion type. The averages of both distributions remained the same and the two sample distributions were statistically different.

Process Improvement One potential use for the adhesion test is for process improvement. During processing the wires

are coated in a reel-to-reel continuous process. After coating, the wires go through a straightening

process and subsequent burn-in step. The process steps are believed to reduce the adhesion of the

Teflon coating. If the new test method was capable of showing a difference in adhesion, steps could be

taken to improve the production process and the test could then be used to determine success or failure

of this process alteration. The first step was to test the same batch of wire in three conditions: as-

coated, as-straightened, and post burn-in to determine if there was a significant change in coating

adhesion after each processing step. The wire tested was the well adhered wire that has been used in all

previous testing.

The wires were first tested using the microscope to determine if there were visible differences

in the coating after each processing step. Twenty wires at each condition were then tested using

electrical continuity as a method of failure detection. During testing, failure at the edge of the wire

where the cylinder made contact was ignored; failure weight was recorded when the coating failed near

the center of the wire sample length.

2.051.851.651.451.25

9

8

7

6

5

4

3

2

1

0

Failure Weight (lb)

Fre

qu

en

cy

Poorly Adhered

Well Adhered

Variable


22

Adhesion Testing

Failure Morphology

The coatings showed a difference in failure morphology after each processing step (Figure 22).

Similar to previous tests of the well adhered wire, the coating in the as-coated condition showed distinct

lines on either side of the testing area due to the good adhesion preventing the Teflon from being

pushed aside during testing. The coating in the as-straightened condition showed a rough, ragged edge

on either side of the testing area, similar to the poorly adhered coating in previous testing. The post

burn-in coating also showed a rough edge; however the edge appeared to be better adhered than the

as-straightened condition.

Figure 22- The wires show different failures at each processing step: as-coated (a), as-straightened (b) and post burn-in (c). The adhesion differences are shown in the edges of the testing area; the good adhesion of the as-coated wire has distinct edges while the next processing steps have rough edges due to poorer adhesion.

0.01 in

0.01 in

0.01 in

Smooth edges

Rough edges

Rough edges

(a)

(c)

(b)

23

Electrical Continuity

The adhesion differences were also apparent during testing with the voltmeter (Figure 23). The

average adhesion of the as-coated wires was 2.1 lb; the average dropped to 1.56 lb for the as-

straightened and rose to 1.7 lb for the post burn-in wires. The straightening process has shown to

decrease the adhesion of the Teflon due to the stress involved in straightening. The burn-in process

seems to have increased the adhesion relative to the straightened adhesion level, possibly due to the

heat relieving some stress on the coating.

Figure 23- A comparison of the well adhered wires after each processing step. The adhesion level dropped after straightening and increased after burn-in.

The contact cylinder was visibly altered during testing. A discoloration could be seen on the area

of the cylinder that contacted the wire. When the discoloration became apparent the cylinder was

rotated to a new position to allow a fresh section for testing. After testing, the residuals were plotted

against the observation order to determine if the discoloration was evidence of a process drift (Figure

24). Each time the cylinder was rotated the residuals dropped, and there was an increase in the

residuals over time, indicating that as the cylinder was being affected by the testing.

Post burn-inAs-straightenedAs-coated

2.50

2.25

2.00

1.75

1.50

Failu

re W

eigh

t (l

b)

24

Figure 24- The residuals were plotted against the order in which the tests were done to determine if there was a process drift, as suggested by the discoloration of the cylinder. Each time the cylinder was rotated corresponds with a low residual and a successive increase in residuals over time. The vertical lines denote each rotation of the cylinder.

The process drift could have been caused by Teflon build-up on the cylinder or the cylinder

being worn during testing. After the discoloration was seen, the cylinder was cleaned and testing

continued. There was no drop in residuals after the cylinder was cleaned; however, after the cylinder

was rotated there was a drop indicating that the cylinder was being worn rather than being affected by

Teflon build-up. The wear on the cylinder has not occurred before during previous testing, and was likely

due to the change in testing method to discount edge effects. Instead of stopping the test after the first

bare wire was exposed, the test continued, causing the contact cylinder to be dragged along stainless

steel rather than Teflon. The steel on steel contact appeared to have damaged the contact cylinder.

Spray Coated Wires An alternative to the reel-to-reel processing method is a spray coating method. The new test

method could be used to determine a difference in adhesion between the spray coated and reel-to-reel

coated wires. Twenty spray coated wires were tested using the same testing method and evaluated

605550454035302520151051

0.3

0.2

0.1

0.0

-0.1

-0.2

-0.3

Observation Order

Res

idu

al

25

using the microscope and electrical continuity. The wires had been through all processing steps and

sterilization.

Adhesion Testing

The microscope showed a difference in the failure morphology of the spray coated wires

compared to the reel-to-reel coated wire (Figure 25). The spray coated wire tended to fail in small

sections rather than along the entire wire.

Figure 25- Failure of the spray coated wire at 1.25 lb. The coating fails in sections rather than failing continuously along the wire.

The spray coated wires failed at an average of 1.35 lb, lower than the averages for the reel-to-

reel coated wires (Figure 26). The average failure weights may not be directly comparable to the reel-to-

reel coated wires because the failure modes were different, and the voltmeter determined failure to be

when the small sections of the coating failed. Further testing could be done to determine when the

coating failed along the entire wire, making the test more comparable to previous tests.

0.01 in

Failed sections

26

Figure 26- Results from testing 20 spray coated wires. The average was lower than all previous tests; however, the failure morphology is not directly comparable to reel to reel coated wires.

Conclusions and Recommendations

The prototype and test method were successful in determining a difference in Teflon adhesion

of the two types of wires with a p-value of less than 0.001 when using either a microscope or

electrical continuity for failure detection.

The prototype showed statistical differences in adhesion at different processing steps, showing

potential to be used for process improvement.

With some alterations the prototype can be used to determine differences in adhesion between

different coating methods.

Recommendations include automating the test and regulating slide speed to reduce operator

variance and replacing the contact cylinder with a harder material to prevent process drift due

to wear of the cylinder.

1.751.651.551.451.351.251.15

6

5

4

3

2

1

0

Failure Weight (lb)

Fre

qu

en

cy

Spray Coated S = 0.165 X = 1.34

27

References 1. Healthwise Staff. "Coronary Artery Disease." Myhealth.alberta, 5 June 2012. Web. 21 May

2013."Clogged Arteries (Arterial Plaque)." WebMD. WebMD, 2005. Web. 11 Dec. 2012.

2. "Coronary Heart Disease." Health Guide. New York Times, 2012. Web. 11 Dec. 2012.

3. "A Patient's Guide to Heart Surgery." Usc.edu. University of Southern California Keck School of

Medicine, n.d. Web. 13 Dec. 2012.

4. "Percutaneous Transluminal Coronary Angioplasty/Stent Placement." Stanford Hospital and

Clinics. Stanford Medicine, 2012. Web. 12 Dec. 2012.

5. "Interventional Cardiology." Central Sydney Cardiology. N.p., 2012. Web. 14 Dec. 2012.

6. Erglis, Andrejs, Inga Narbute, and Dace Sondore. "Tools & Techniques: Coronary Guide

Wires." EuroIntervention (2010): n. pag. Pcronline. Web. 13 Dec. 2012.

7. Gibson, Michael. "Guidewire." Wikidoc. Abbott Vascular, 24 Oct. 2011. Web. 14 Dec. 2012.

8. Shah, Vikram. "Fluoropolymers Selection For Corrosive Media Applications." Antico Pumps RSS.

Centrizone, 3 May 2010. Web. 14 Dec. 2012.

9. Funderburg, Anne C. "Making Teflon Stick." Editorial. Invention and Technology Magazine2000:

n. pag. Whitetrout.com. 26 Oct. 2012. Web. 18 Oct. 2012.

10. Marotta, Christine, Mike Williams, and Nicole Laput. "Adhesives for the Assembly of Hard to

Bind Plastics." Henkelna.com. Henkel, n.d. Web. 18 Oct. 2012.

11. Schröder, J. "The Mechanical Properties of Guidewires. Part III: Sliding Friction."Cardiovasc

Intervent Radiol, 16.2: 93-97.

12. “ASTM D4541-09 Standard Test Method for Pull-off Strength of Coatings Using Portable

Adhesion Testers.” astm.org. ASTM International, n.d. Web. 10 November 2012.

13. “ASTM D903-98 Standard Test Method for Peel or Stripping Strength of Adhesive Bonds.”

astm.org. ASTM International, n.d. Web. 10 November 2012.

14. “ASTM D2197-10 Standard Test Method for Adhesion of Organic Coatings by Scrape Adhesion.”

astm.org. ASTM International, n.d. Web. 17 September 2012.

Documents

Development of a Production Line Appropriate Test for